Standing/walking stabilization of humanoid robot by visual servoing concept through online-visual-pose-estimation

Although many papers have been published on visual tracking, visual servoing of the vision-based robotics, there are only few research study on using vision to improve the standing and walking stabilization for legged robots. In this paper, we set two cameras as the humanoid's eyes, to observe a static object in front, and using the visual measured information to control the robot keeping a desired head-top's position/orientation, in order to help the robot to prevent from unstable motion, such as the falling down action caused by gravity or dangerous slipping motion happened suddenly.